***************************************************************************
SOO is published by Gahmken Press (P.O. Box 1467, Newport, Oregon 97365), is numbered consecutively as independent monographs, and appear at irregular intervals; there are no subscriptions. SOO is not affiliated with nor supported by "Oregon Fund for Ornithology" (1988 Oregon Birds 14[4]:315) or Oregon Field Ornithologists. Copies are donated to several university and research libraries.
---------------------------------------------------------------------------
Library of Congress Cataloguing-in-Publication Data
Bayer, Range D., 1947-
Cormorant harassment to protect juvenile salmonids in Tillamook
County, Oregon / by
Range D. Bayer.
p. cm. -- (Studies in Oregon ornithology; no. 9)
Includes bibliographical references (p. ).
ISBN 0-939819-09-0 (pbk.)
1. Fishery conservation--Oregon--Tillamook County. 2. Cormorants-
-Behavior--Oregon--Tillamook County. 3. Coho salmon--Oregon--Tillamook
County. 4. Steelhead (Fish)--Oregon--Tillamook County. I. Title. II.
Series.
SH222.O7 B38 2000
333.95'616'0979544--dc21 00-023174
---------------------------------------------------------------------------
Copyright © 2000 by Range (Richard) D. Bayer. Without charge, permission is freely given to anyone to use any means to copy this publication as long as it is credited as the source.
This document should be cited as:
Bayer, R. D. 2000. Cormorant Harassment to Protect Juvenile Salmonids in Tillamook County, Oregon. Studies in Oregon Ornithology No. 9. 66 pages.
***************************************************************************
The goal of the Oregon Plan is to restore wild coho and wild steelhead runs. Under the federal Endangered Species Act, wild coho salmon along the Oregon Coast are listed as Threatened and wild Oregon Coast steelhead are a candidate for listing. Although cormorants have been hazed at the Nehalem Estuary for at least 10 years and at the Tillamook and Nestucca Estuaries for at least three years, spawning ground counts of wild coho salmon, winter steelhead, and fall chinook have averaged less since hazing began. Thus, hazing does not appear to be useful in recovering wild salmonids.
Hazing is not correlated with consistently improved hatchery returns. The survival of Coded Wire Tag marked coho smolts at the Nehalem was about the same whether hazing occurred or not, the percent return for coho smolts was not significantly greater at the hazed Nehalem than at the nonhazed Salmon River, and the number of returning adult coho salmon was significantly greater with hazing at the Nehalem hatchery but not at the Trask hatchery in the Tillamook Basin. For winter steelhead, the number of returning adults to the Nehalem and jacks to the Cedar Creek hatchery in the Nestucca Basin did not increase significantly with hazing, but the number of jacks returning to the Nehalem did.
Changes in fisheries subsequent to hazing are mixed. Coho catches increased with hazing at the Nehalem but not at the Tillamook Basin. Nehalem steelhead catches averaged less with hazing, but chinook fisheries have grown. However, the increase in chinook catches occurred as the number of wild chinook at spawning areas declined, so the larger catch may be a consequence of a greater harvest of wild chinook rather than hazing.
Returns may not have increased with hazing because it was ineffective in substantially reducing predation, because smolts saved by hazing died anyway, or because other factors such as unfavorable ocean conditions may have been much more important in affecting smolt survival than hazing.
In any case, hazing does not appear to be a panacea for salmonid recovery, and it has costs. During 1996-1999, the Oregon Legislature spent $100,000 for cormorant hazing, and a biological cost of hazing is the disturbance of wildlife other than cormorants.
***************************************************************************
ADC Animal Damage Control Program of U.S. Dept. of Agriculture,
which is currently known as the Wildlife Services Agency
of the U.S. Dept. of Agriculture
CWT Coded Wire Tag
ESA federal Endangered Species Act
NMFS National Marine Fisheries Service
OAR Oregon Administrative Rules
OCSRI Oregon Coastal Salmon Restoration Initiative (1997a,b)
ODFW Oregon Department of Fish and Wildlife
OSU Oregon State University
PDT Pacific Daylight Time
PFMC Pacific Fishery Management Council
SPP Smolt Protection Program, which is conducted by private
individuals, under contract to the ODFW (see section C-1)
STEP Salmon and Trout Enhancement Program of ODFW
USFWS U.S. Fish and Wildlife Service
***************************************************************************
I thank Tom Stahl (OSU) and Bill Pearcy (OSU) for their time and constructive comments in reviewing an earlier draft of this paper during the summer of 1999 and Bob Buckman (ODFW) for his patient responses to my questions about salmonid biology.
All salmonid data in this report are from ODFW reports or were generously provided upon request from several ODFW staff. I am grateful to the following individuals, listed alphabetically, for providing information or data since 1988: Kay Brown (ODFW), Bob Buckman (ODFW), Tracy Cabe (ODFW), Mark Chilcote (ODFW), Kathleen Confer, T. Edwin Cummings (ODFW), Bill Haight (ODFW), Thomas R. Hoffman (ADC), Steve Jacobs (ODFW), Steve Johnson (ODFW), Ken Kenaston (ODFW), Rick Klumph (ODFW), John Leppink (ODFW), Mark Lewis (ODFW), Roy Lowe (USFWS), Laimons Osis (ODFW), Thomas Riley (USFWS), Leslie Schaeffer (ODFW), Eric Schindler (ODFW), Mario Solazzi (ODFW), Doug Taylor (ODFW), Jo Walin, Don Watson, Ron Williams (ODFW), and Gary Yeager (ODFW).
I thank Janet Webster, Susan Gilmont, and Judy Mullen of the Guin Library at the OSU Hatfield Marine Science Center for assistance in obtaining references.
***************************************************************************
Executive Summary_______________________________________________________ 3
Acronyms________________________________________________________________ 3
Acknowledgments_________________________________________________________ 4
A. Introduction________________________________________________________ 5
B. Study Areas_________________________________________________________ 5
C. Hazing Methods______________________________________________________ 7
D. Methods of Counting Cormorants in 1996-1998_________________________11
E. Methods of Correlating Hazing with Salmonid Returns_________________11
F. Results and Discussion: Smolt Predators and Hazing__________________15
G. Results and Discussion: Wild Salmonid Abundance_____________________20
H. Results and Discussion: Coho CWT Returns____________________________32
I. Results and Discussion: Hatchery Returns____________________________34
J. Results and Discussion: Fisheries Catches___________________________41
K. Concluding Remarks__________________________________________________50
Appendix I. Common and scientific names of animals_____________________54
Appendix II. Coho, steelhead, and chinook life history information_____55
Appendix III. Number of coho and winter steelhead smolts released at
Nehalem____________________________________________________________57
Appendix IV. Lack of rigorous controls in testing the effects of
hazing_____________________________________________________________58
Appendix V. Factors other than hazing that can affect salmonid
returns____________________________________________________________59
Literature Cited________________________________________________________62
++++ The last page of the printed version of SOO No. 9 is p. 66.
Tables and Figures
Table-Page No. Figure-Page No.
Table 1-7 Figure 1-6
Table 2-9 Figure 2-21
Table 3-10 Figure 3-22
Table 4-13 Figure 4-24
Table 5-17 Figure 5-25
Table 6-18 Figure 6-26
Table 7-19 Figure 7-28
Table 8-20 Figure 8-29
Table 9-24 Figure 9-30
Table 10-27 Figure 10-31
Table 11-30 Figure 11-33
Table 12-32 Figure 12-34
Table 13-36 Figure 13-35
Table 14-42 Figure 14-37
Table 15-46 Figure 15-39
Table 16-52 Figure 16-40
Table 17-57 Figure 17-43
Table 18-58 Figure 18-44
Figure 19-47
Figure 20-48
Figure 21-50
Figure 22-61
***************************************************************************
The purpose of the Oregon Coastal Salmon Restoration Initiative (OCSRI 1997a,b) is to recover populations of wild coastal coho salmon and wild coastal steelhead (common and scientific names are given in Appendix I). After the OCRSI was proposed, wild Oregon coastal coho salmon were listed as Threatened under the federal Endangered Species Act (ESA)(National Marine Fisheries Service [NMFS] 1998), and wild Oregon coastal steelhead are a candidate for listing under the federal ESA (Busby et al. 1996, Chilcote 1998; http://www.nwr.noaa.gov as of 12/11/1999). The OCRSI has been extended in the Oregon Plan to all at- risk wild salmonids throughout the state (Kitzhaber 1999). Thus, the primary state and federal goal is the recovery of wild salmonid populations.
Accordingly, actions that would increase the number of wild adult coho or steelhead to Oregon Coast spawning areas would be beneficial. Such actions could include cormorant harassment (hazing), which has been claimed in testimony before the Oregon Legislature and in newspaper and magazine articles to improve salmonid returns (Erickson 1989a,b,d; 1992, 1993, 1995a,b; Monroe 1995b, 1996a; Nokes 1995). Acting on these assertions, the Legislature has allowed and funded harassment (section C-1).
This paper has two purposes. First, to determine if cormorant hazing in Tillamook County is correlated with increased returns of wild salmonids, and thus possibly important in achieving the goal of the Oregon Plan and ESA. The second purpose is to see if hatchery salmonid returns and fishery catches have been enhanced with hazing.
***************************************************************************
Information about the estuarine area, drainage basin size, and average freshwater flow in April-June (when many coho and winter steelhead smolts migrate through tidewater; Appendix II) for estuaries included in this report are in Table 1, and their location is shown in Fig. 1. Note that estuaries differ markedly in these characteristics, and annual variation is to be expected.
Cormorant hazing has occurred in Tillamook County in tidewater portions of the Nehalem, Tillamook, and Nestucca basins (section C). Each basin (Fig. 1) has one hatchery: the North Fork of the Nehalem hatchery in the Nehalem Basin, the Trask hatchery in the Tillamook Basin, and the Cedar Creek hatchery in the Nestucca Basin.
The Salmon River, Siletz, and Alsea basins are most often used for comparison with the Nehalem (where most hazing has occurred, section C) because they are the nearest basins that are most like the Nehalem. For example, like the Nehalem, many hatchery coho and winter steelhead smolts have been released into each (Kostow 1995:Appendix A; Lewis 1997) and their sport fisheries are large enough to compare to the Nehalem (ODFW 1998b). However, they are 54-90 mi south of the Nehalem (Table 1), so human and environmental factors may affect them differently than the Nehalem. Other nonhazed basins (e.g., the Necanicum) are also included if spawning ground counts are available or if their coho or winter steelhead sports fisheries are substantial (ODFW 1998b). The Columbia is excluded because it has an estuary area of 147 sq mi (ODSL 1973), which is more than twice as large as all the rest of these estuaries combined, and it also differs politically and biologically from other coastal estuaries.
Compilations of information about salmonids in coastal basins exist for the Tillamook (Ellis 1998), Yaquina, Alsea, Salmon, Siletz, and other mid-coast stream basins (ODFW 1991,1997a-e); but fisheries management plans have not yet been prepared for the Nehalem, Tillamook, or Nestucca Basins (Rick Klumph, ODFW, pers. comm.).
---------------------------------------------------------------------------FIGURE 1. Location of bays/estuaries (east of shoreline) that are mentioned in the text, with nearby large towns or landmarks listed west of the shoreline. The shoreline of Tillamook County is approximately from Cape Falcon to Cascade Head.
--------------------------------------------------------------------------TABLE 1. Oregon estuaries included in analyses of cormorant hazing. ?=unknown.
--------------------------------------------------------------------------
Mean Distance &
Freshwater Direction
Estuary Drainage Inflow @@ from Known
Area Basin*@ (1,000 cfs) Nehalem** Corm.
Estuary (sq mi)* (sq mi) APR MAY JUN (mi) Hazing
-----------------------------------------------------------------------
Necanicum <1 87 ? ? ? 21 north no
Elk/Ecola Creek # ? ? ? ? ? 13 north no
Nehalem 4 855 3.3 1.6 0.7 0 yes
Tillamook 13 540 3.8 2.1 1.1 9 south yes
Nestucca 2 322 ? ? ? 37 south yes
Salmon River <1 75 ? ? ? 46 south no
Siletz 2 373 2.7 1.5 0.9 54 south no
Yaquina 6 253 1.2 0.5 0.3 76 south no
Beaver Creek # ? ? ? ? ? 83 south no
Alsea 3 474 2.1 1.3 0.6 90 south no
Umpqua 11 4,560 ? ? ? 137 south no
# Estuary so small that it is not included in ODSL (1973), Percy et al.
(1974), or Shirzad et al. (1988).
* ODSL (1973) area in acres converted to sq miles.
*@ Percy et al. (1974:3).
@@ Shirzad et al. (1988); cfs=cubic feet per second.
** Straight line distance to the mouth of each estuary.
***************************************************************************
A state permit is required to legally harass migratory birds on public lands or waters of Oregon (Bayer 1989:10, ODFW 1998a:11, 13, 19, 22). Roy Lowe (U.S. Fish and Wildlife Service [USFWS], pers. comm.) thought that a federal permit might also be required if such harassment directly impacted nesting success.
In the spring of 1988, the Oregon Department of Fish and Wildlife (ODFW) Director, Randy Fisher, authorized the issuance of a series of state permits to allow private citizens to harass cormorants on public waters of Nehalem and Tillamook Bays (Bayer 1989:12). The harassers were not monitored by the ODFW (Bill Haight, ODFW, pers. comm.) or USFWS (Thomas Riley, USFWS Law Enforcement, pers. comm.). The Director can issue such permits by declaring an emergency and assuming the powers of the Oregon Fish and Wildlife Commission.
In 1989, the Oregon Fish and Wildlife Commission and the ODFW Director refused to issue any more permits, although petitioned to do so (e.g., Erickson 1989e, Bayer 1989:13-14, ODFW 1998a:22). Consequently, harassment proponents turned to the Oregon Legislature, and House Bill (HB) 3185 was introduced into the 1989 Session to allow cormorant harassment, but it did not pass (Bayer 1989:13-14, 55-58; ODFW 1998a:22). HB 2735 passed during the 1989 Session and directed the ODFW to study how to protect hatchery smolts; this resulted in Schaeffer's (1991) thorough literature review.
In 1995-1997, the Oregon Legislature passed bills directing the Commission to issue not more than three permits to private citizens to harass cormorants on public waters in Tillamook County and to allocate $25,000 of state funds per year to compensate the permittee's expenses in conducting cormorant harassment (1995 Senate Bill 707, 1996 [Special Session] House Bill 3483, and 1997 Senate Bill 5503; see http://www.leg.state.or.us). The 1995 bill was vetoed by Governor Kitzhaber (Monroe 1995b); however, the 1996 bill passed, so hazing is now authorized each year. But state funding of hazing must be approved each biennium by the Oregon Legislature; funding in the amount of $50,000/biennium was approved for 1996-1999. To obtain these funds, an annual report for hazing in these Smolt Protection Programs (SPP) is required for each bay; these reports are available from the ODFW (SPP 1996-1998). The permittees were not monitored by the ODFW (Kay Brown, ODFW, pers. comm.).
In 1997, Oregon Senate Bill 622 (see http://www.leg.state.or.us) was introduced that would have expanded cormorant harassment to all Oregon coastal streams and the main stem of the Columbia River and also would have removed a limit on the number of permits issued. A work session and public hearing were held, but it was not passed out of committee.
In 1999, the Legislature funded hazing for 2000-2001 (Charlie Bruce, ODFW, pers. comm.).
C-2a. INTRODUCTION. Hazing methods need to be clearly stated, so that their effectiveness can be evaluated or others can use them in planning their own hazing program. Unfortunately, hazing methods in the SPP have not been adequately detailed.
C-2b. NEHALEM. Hazing was by cracker shells in 1988 and by a fast boat in following years (Table 2). Fast boats scared away cormorants by moving back and forth over schools of juvenile salmonids, while sometimes playing cormorant distress calls (Monroe 1995b, 1996a,b).
Scarecrows were also placed where smolts seemed to concentrate but seemed ineffective in keeping cormorants away after a day or so (SPP 1997-1998). A "Wytech" device (SPP 1996: April 18 and May 2-6, 16, 20) may have also been used (e.g., "Wytech seems to make birds nervous but very limited range. Will call and request long range unit"), but it is not explained what it is. Additionally, Monroe (1995b) wrote:
"Smolt protectors tried a remote-control airplane one spring, but one of the operators wrecked it on the corner of a winding highway one afternoon while chasing a neighbor's truck. Volunteers went back to their boats."
C-2c. TILLAMOOK. In 1988, harassment was by cracker or screamer shells fired from firearms (Bayer 1989:12). In 1996 and 1998, the method is not given in SPP reports but was probably by using a fast boat like at the Nehalem. In 1997, a boat was used to patrol the upper and lower bay, and someone ran a hovercraft over Kilchis Flats during low tide on April 25 that disturbed cormorants in water too shallow for the hazing boat (SPP 1997). In 1997, a scarecrow at the mouth of the Wilson River was also used to deter cormorants (SPP 1997).
C-2d. NESTUCCA. Presumably, a boat was used to chase cormorants, and the use of a mannequin (scarecrow) or "dummy" was also mentioned (SPP 1996-1998).
C-3a. NEHALEM. It is unclear if hazing occurred during 1985-1987. In 1985, hazing supplies were issued by the U.S. Dept. of Agriculture Animal Damage Control Program (ADC) to the President of the Nehalem Salmon and Trout Enhancement Program (STEP) group to be used by STEP volunteers to harass cormorants; these supplies may have included 12 gauge shotgun cracker shells and/or audio distress calls (Bayer 1989:11). In 1986, the ADC were contacted about hazing by staff from the ODFW North Fork Nehalem fish hatchery (Hoffman and Hall 1988). Hazing may have occurred in 1987 because Nokes (1995) indicated that there had been eight years of hazing at the Nehalem; since there was no hazing in 1995 (Monroe 1995b), backdating from 1994 suggests that hazing occurred in 1987. In any case, the interest in hazing was sufficient that by 1988, harassers were aware that they had to have permits to do so legally and went through the process of requesting these permits.
Hazing, with or without authorization, was reported during 1988-1994 (Table 2). This hazing was discussed during public testimony in the Oregon Legislature in 1989 (Parks 1989) and 1995 (Erickson 1995b), and there were also articles about it in fishing publications (e.g., Erickson 1992, 1993; Orcutt 1992), in the Oregonian (Oregon's largest newspaper)(Monroe 1995a,b; Nokes 1995), and in the Tillamook County newspaper (e.g., Erickson 1995a). Further, Roy Lowe (USFWS biologist, pers. comm.) notes that Erickson discussed his 1989-1992 Nehalem Bay hazing activities at the salmon predator workshop chaired by the Coastal Oregon Marine Experiment Station at the OSU Hatfield Marine Science Center on 28 September 1992, copies of Erickson's (1992) article about his hazing were available at the December 1992 Oregon Governor's Coastal Salmonid Restoration Initiative meeting at Otter Crest, and Lowe wrote the ODFW Director Randy Fisher about these unauthorized activities on 5 January 1993.
There was no hazing in 1995 (Monroe 1995b). This lapse may have been because hazers were no longer compensated for their expenses; Erickson (1995b) testified that a private individual had previously paid for the costs of gas and oil for hazing at the Nehalem, which was about $3,000 per year and that hazers had volunteered their time.
C-3b. TILLAMOOK AND NESTUCCA. State permits were given to allow hazing at Tillamook Bay in 1988 (section C-1). In 1993-1994, the boat used for hazing at Nehalem Bay was tested at Tillamook and Nestucca Bays during three days to see if it could haze cormorants (Erickson 1995b), but this test hazing would have been too brief to have a significant effect on reducing cormorant predation. In 1996-1998, hazing at Tillamook and Nestucca Bays occurred as part of the SPP (1996-1998).
C-4a. NEHALEM. When documented, hazing occurred in spring from early or late April through early to mid-June (Table 2). In 1996 and 1998 SPP reports, there were notes for each day during this period, but it is not clear if hazing occurred daily; however, in 1996, there were at least three days without hazing and there were other days in which hazing probably did not occur because of bad weather. In the 1997 SPP report, the contractor was said to have hazed cormorants every day during the season.
Hatchery smolt releases occurred before hazing was authorized in at least 1996-1998 (Table 3), so cormorants may have been feeding on smolts prior to hazing. However, hazing occurred prior to authorization in at least 1998 (Monroe 1998).
C-4b. TILLAMOOK. Hazing was reported to have occurred during 21 April-20 June 1988, 15 April-7 June 1996, 8 April-15 June 1997, and 27 April-15 June 1998 (Bayer 1989:12, SPP 1996-1998), but hazing may also have occurred prior to authorization (Monroe 1998). In SPP (1996), there are daily records of the numbers of cormorants, but it is not clear if hazing occurred each day. In 1997 and 1998, there are no records of hazing during 12 and eight days of each season, respectively; six of these missed days in 1997 and one of the missed days in 1998 were before May 15.
C-4c. NESTUCCA. Harassment was recorded as starting on April 15, April 13, and April 14 in 1996-1998, respectively, and ended on June 15 each year (SPP 1996-1998). In SPP (1996), there are daily records, but it is not clear if hazing occurred each day. In 1997, there are no records during two days of the season, and, in 1998, there were no days without notes (SPP 1997-1998).
---------------------------------------------------------------------------TABLE 2. Authorization, method, and duration of cormorant harassment at Nehalem during 1988-1998. The hazing dates are those given on permits, but hazing may have occurred earlier, as in 1998 (Monroe 1998). It is unclear if hazing occurred in 1985-1987. corm.=cormorants.
---------------------------------------------------------------------------
Hazing Legal
Permit Species Hazing....................
Yr Issued to Haze Method Duration Reference(s)
---------------------------------------------------------------------------
88 yes corm. shell A 4/8-6/6 B
89 no none boat spring C
90 no none boat spring D
91 no none boat spring D
92 no none boat spring D
93 no none boat spring D
94 no none boat spring D
95 no none no hazing no hazing Monroe 1995b
96 yes corm. boat & scarecrow 4/8-6/18 Monroe 1996a,b; SPP 1996
97 yes corm. boat & scarecrow 4/7-6/5 SPP 1997
98 yes corm. boat & scarecrow 4/24?-6/15 E
A Cracker or screamer shells shot from firearms.
B Hendrickson 1988, McAllister 1988, Bayer 1989:12, Erickson 1989d,
1992, 1993, 1995b; Monroe 1995b, Nokes 1995.
C Parks 1989, Bayer 1989:13, Erickson 1992, 1993, 1995b; Monroe 1995b,
Nokes 1995.
D Erickson 1992, 1993, 1995a,b; Monroe 1995b, Nokes 1995.
E In 1998, authorization started on April 24, but hazing had already
started (SPP 1998, Monroe 1998).
-------------------------------------------------------------------------
---------------------------------------------------------------------------TABLE 3. Number of coho and winter steelhead fry, fingerlings, or smolts released into the Nehalem Basin during 1988-1998. These data were provided by John Leppink and Tracy Cabe of the ODFW. Dates of hazing are from Table 2. Hazing dates are from permits, but hazing in at least 1998 occurred before permits were issued (Monroe 1998). Thousands=thousands of fish released, Fr=fry, Fn=fingerling, Sm=smolt, WSte=winter steelhead. *=released at the North Fork of Nehalem hatchery; @=released elsewhere in the Nehalem Basin, *@=released both at the hatchery and also elsewhere in the Nehalem Basin on the same day.
---------------------------------------------------------------------------
Juv Salmonid Release Juv Salmonid Release
Thou- Hazing Thou- Hazing
Yr Date sands Age Kind Dates Yr Date sands Age Kind Dates
------------------------------------ -------------------------------------
88 3/2 67 Sm* Coho 93 2/8 65 Sm* Coho
3/23 186 Sm* Coho 2/19 65 Sm* Coho
3/28 487 Sm* Coho 3/3 65 Sm* Coho
4/4 104 Sm*@ WSte 4/13 565 Sm* Coho ?
4/5 53 Sm*@ WSte 4/8-6/6 4/13 42 Sm* WSte
6/8 124 Fn* Coho 4/15 48 Sm@ WSte
4/16 35 Sm* WSte
89 1/31 206 Sm* Coho 9/29 26 Fn@ WSte
4/14 599 Sm* Coho ?
4/14 31 Sm* WSte 94 2/28 213 Sm* Coho
4/19 123 Sm*@ WSte 4/15 627 Sm* Coho ?
4/28 53 Fn* Coho 4/15 82 Sm* WSte
6/26 24 Fr* WSte 4/27 51 Sm@ WSte
6/29 24 Fr* WSte 4/28 12 Sm* WSte
10/20 8 Fn* WSte
10/26 4 Fn* WSte 95 1/30 214 Sm* Coho
4/10 7 Sm@ WSte none
90 2/2 65 Sm* Coho 4/14 576 Sm* Coho
2/12 66 Sm* Coho 4/14 80 Sm* WSte
2/22 67 Sm* Coho
4/2 55 Sm*@ WSte 96 3/15 137 Sm* Coho 4/8-6/18
4/3 102 Sm*@ WSte ? 4/9 7 Sm@ WSte
4/16 632 Sm* Coho 4/12 83 Sm* WSte
4/30 117 Fn* Coho 4/15 500 Sm* Coho
6/20 2 Fr* WSte
7/10 6 Fr* WSte 97 3/14 140 Sm* Coho
10/30 6 Fn@ WSte 4/2 15 Sm@ WSte 4/7-6/5
4/15 489 Sm* Coho
91 2/25 57 Sm* Coho 4/15 78 Sm* WSte
3/4 115 Sm* Coho
4/2 20 Sm@ WSte 98 3/31 103 Sm* Coho
4/3 83 Sm*@ WSte ? 4/13 15 Sm@ WSte 4/24?-6/15
4/4 7 Sm@ WSte 4/30 90 Sm* Coho
4/7 30 Sm* WSte 4/30 63 Sm* WSte
4/11 564 Sm* Coho 5/12 49 Fn@ Coho
92 1/24 69 Sm* Coho
2/3 67 Sm* Coho
2/14 70 Sm* Coho
3/24 11 Sm* WSte
3/30 63 Sm*@ WSte
3/31 70 Sm*@ WSte
4/13 626 Sm* Coho ?
4/20 54 Fn@ Coho
10/9 13 Sm@ WSte A
A 10/9/92 steelhead "smolts" weighed 13.8/lb, but smolts in spring weigh
5-7/lb, so these "smolts" were not as large as those in spring.
---------------------------------------------------------------------------
C-5a. NEHALEM. During 1988-1994, it is unknown when hazing was done during the day. In 1996, hazing is recorded as occurring from about 7 AM Pacific Daylight Time (PDT) until early afternoon on some days; on other days, it is stated that hazing was done in the morning, but it is unclear if it was also done in the afternoon (SPP 1996). In 1997, it is stated that hazing occurred every day from 7 AM until 3 PM PDT (SPP 1997). In 1998, the daily duration of hazing is not specified (SPP 1998).
C-5b. TILLAMOOK. The duration of hazing during a day is not clearly recorded in 1996 and 1998 (SPP 1996, 1998). In 1997, the duration is not explicitly stated, but reports of hazing usually ranged from about 7 AM to 2 PM PDT; however, hazing in late May and early June was often only recorded in the afternoon and/or evening (SPP 1997).
C-5c. NESTUCCA. In 1996, the time of hazing was rarely recorded, but on three days when it was, patrolling occurred 11-14 hr/day (SPP 1996). In 1997, hazing was recorded as ending at 6:30-8 PM PDT on twelve days, but neither the starting time nor the ending time for other days was recorded (SPP 1997). In 1998, the starting and ending times were generally given, and hazing duration was usually about 11-14 hr/day (SPP 1998).
***************************************************************************
Descriptions of the methods of counting cormorants are essential but are insufficiently detailed in SPP (1996-1998). Examples of critical information include the starting and ending times of counts, the power of binoculars or telescope used for censusing, the area included in counts, the sites where censuses were made from, the methods of avoiding counting the same cormorants more than once during a census, and whether cormorant numbers were counted or estimated.
The census techniques that are described in SPP (1996-1998) raise questions about the accuracy of their cormorant counts for several reasons. First, it is essential that it is stated what bird species is being counted, but many counts (e.g., Nehalem Bay in 1996 and Nestucca Bay in 1996-1998) only give a number of "birds." Since only cormorants are supposed to be hazed, the reader could guess that the counts are of cormorants; however, other species were also disturbed (section F-3). Second, counts need to be of the entire area of concern, but only counts at Nehalem in 1998 state that they were for the "entire estuary"; other SPP counts appear to be point counts for where the hazer was at a particular time, not the whole bay. Third, counts need to be throughout the day to determine how cormorant abundance may change, but the 1998 Nehalem counts were just at 7 or 8 AM. Finally, a census should not include cormorants that have already been counted during that census. This can be difficult, especially if they are moving as they would be if they are being hazed, and it is not stated in any report how recounting the same birds was avoided.
Similar problems with cormorant counts in the 1999 SPP report are also described in Stahl et al. (2000:33-34).
***************************************************************************
Coho and steelhead are examined because hazing has been claimed to have greatly improved their returns (Erickson 1989a,b,d; 1993, 1995a; Monroe 1995b, 1996a; Nokes 1995). Further, smolts of both species are released (Table 3) or migrating to the ocean during spring (Appendix II-1 and II-2c), when hazing occurred. Chinook are included because juvenile chinook were observed at Nestucca Bay during early June 1996 hazing (SPP 1996) and may have been present at each bay during the time of hazing (Appendix II-3b), so hazing may have also affected their survival.
E-2a. INTRODUCTION. Stahl et al. (2000) used radiotracking to study the survival of hatchery coho smolts between release and their arrival at Fishery Point in Nehalem Estuary during the spring of 1999. Their short-term survival results suggest that there was substantial predation even with hazing and are discussed in section K-7. But adult return data are also essential in determining if hazing may be an important management tool because if hazing does not increase the number of adults returning to spawn or caught in fisheries, then its value is questionable, even if it may increase the survival of smolts migrating through an estuary (section K-3). Although jack and adult returns are affected by factors such as fisheries regulations and environmental conditions as well as hazing (Appendix V), proponents have claimed that hazing has improved jack and adult returns (Erickson 1989a,b,d; 1993, 1995a; Monroe 1995b, 1996a; Nokes 1995). Thus, correlating returns with hazing is appropriate.
E-2b. TIME DELAY OF RETURNS. Based on coho life history (Appendix II-1), any positive effects of hazing in improving survival of outmigrating smolts in the spring of one year may be correlated to increased jack returns during the fall of the same year or adult returns in the fall of the next calendar year. For example, hazing during the spring of 1988 may have affected returning jacks in the fall of 1988 and adults in the fall of 1989.
For winter steelhead, hazing in the spring of one year may affect jack (1-salt) returns after one summer in the ocean and adult returns after one or more additional summers in the ocean (Appendix II-2d). Because the percentage of adults coming back at a particular age can be variable (e.g., an average of 66-80% of adults returned after two summers in the ocean as 2-salts; Appendix II-2d), testing for the effects of hazing would be most robust for adults returning after at least two years and preferably 3-4 yr of hazing; this would be particularly true for wild adults because a higher percentage return as 3-salts and repeat spawners (Appendix II-2d).
Positive effects of hazing in the spring of one year on the survival of juvenile chinook may be correlated to jack returns during the following year and the returns of adults 2-6 years after hazing, when adults were 3-7 yrs old (Appendix II-3c). For example, hazing in the spring of 1988 may affect the returns of chinook jacks in 1989 and of adults mostly during 1990-1993. This large range and the yearly variation in the age of returning males or females (Appendix II-3c) make data analyses for determining the effects of hazing on chinook difficult. Hazing must occur for six consecutive years before it can be certain that all returning adults could have been affected by hazing when they migrated to the ocean as juveniles. The only bay where hazing has occurred that long is Nehalem.
E-3a. SPAWNING GROUND COUNTS. The only long-term measure of wild coho, steelhead, and chinook abundance along the Oregon Coast is spawning ground surveys (Weber and Knispel 1977, Jacobs and Cooney 1997, Chilcote 1998). Although these surveys are only for portions of streams and consequently may not accurately estimate the total number of spawners for unsurveyed areas (Weber and Knispel 1977:52-53, Ellis 1998:3-6, Botkin et al. 1995 cited in Ellis 1998), these surveys are done consistently at the same areas, so these counts can be examined to see if changes in abundance are correlated with hazing.
E-3b. CODED WIRE TAGS (CWT) RETURNS. Some hatchery coho and chinook smolts were marked with CWT's for stock assessment (Lewis 1997). CWT recoveries are from ocean fisheries, hatcheries, and spawning ground surveys, but not freshwater fisheries (Lewis 1997:2, 13). For coho, only Nehalem returns are used for a basin with hazing because there are no CWT data available for the Cedar Creek hatchery (Nestucca Basin), and there are only two years of post-hazing data for the Trask or Trask Pond hatcheries (Tillamook Basin)(Lewis 1997). Chinook CWT data are not analyzed because there are no return data for the Nehalem (Lewis 1997) and hazing at the Tillamook Basin and Nestucca Basin only began consistently in 1996 and not all adults affected by this hazing would return until 2002. I did not find any steelhead CWT data.
E-3c. HATCHERY RETURNS. The number of fish returning to a hatchery can be affected by several factors (Appendix V), including the number of coho and steelhead smolts released (which has been variable at the Nehalem; Appendix III) and the season of counting returning fish (which has not been consistent, Table 4). However, correlating hatchery returns of coho and steelhead with hazing is appropriate because they have been reported as increasing as a consequence of hazing (Erickson 1989a,b,d; 1993, 1995a; Monroe 1995b, 1996a; Nokes 1995); for example, Monroe (1996a) wrote:
"Strong hatchery returns of coho salmon and steelhead at the North
Fork Nehalem fish hatchery have defied El Nino and other conditions
blamed for flagging runs in most other streams. The hatchery's
success coincided so closely with the local 'smolt protection
program' that 1995 legislators finally approved this biennium's
test for Nehalem, Tillamook and Nestucca bays."
Chinook hatchery releases into the Nehalem have been limited or discontinued (Nicholas and Hankin 1989: 253, Kostow 1995:A3-A4). Consequently, a maximum of 70 and generally less than a sum of 25 jack and adult chinook returned annually to the North Fork of Nehalem hatchery during 1984-1997 (John Leppink and Tracy Cabe, ODFW, unpubl. data); these are too few fish to test the effects of hazing.
Since hazing at the Tillamook Basin and Nestucca Basin only began consistently in 1996 and adult chinook can return at up to seven years old (Appendix II-3c), only returns of chinook jacks are tested at hatcheries in these basins to see if they have been affected by hazing.
---------------------------------------------------------------------------TABLE 4. Number of winter steelhead jacks counted and season of counting returning steelhead at the North Fork of Nehalem River hatchery during the 1985-1986 and 1988-1989 Run-Years. Data are from Gary Yeager, John Leppink, Leslie Schaeffer, and Rick Klumph (ODFW, pers. comm.). -=counting ceased before any were counted.
---------------------------------------------------------------------------
Steelhead Jacks/ Season of Counting Steelhead Jacks.....
Month Steelhead
1985- 1988- Season....... No. of
1986 1989 Run-Year Begin End Jacks
---------------------------------------------------------------------------
September 0 0 1985-1986 ? 3/17 15
October 0 0 1986-1987 ? 1/27 0
November 0 0 1987-1988 ? 2/8 0
December 0 0 1988-1989 ? 3/6 46
January 5 31 1989-1990 10/20 ? 0
February 7 15 1990-1991 10/5 2/25 16
March 3 - 1991-1992 ? ? 19
1992-1993 ? 3/26 2
SUM 15 46 1993-1994 12/1 2/24 2
1994-1995 11/23 3/17 16
1995-1996 11/17 3/5 30
1996-1997 11/25 3/10 70
---------------------------------------------------------------------------
E-3d. PUNCHCARD DATA. The origin of salmonids caught in the ocean is unknown unless they have been marked with CWT's, but those caught in an estuary or stream probably originated in that basin, although some hatchery-reared salmonids can return to a basin other than the one into which they were released (Jacobs 1988, Lindsey et al. 1993-1995, Quinn 1997). Estuarine and freshwater fishery catch estimates are based on salmon/steelhead tags (punchcards) and are available through 1996 (ODFW 1998b). Although many factors other than hazing may have affected catches (Appendix V), examining catch data to see if it is correlated with hazing is important because maintaining or increasing fisheries is one goal of salmonid management.
E-4a. COMPARING HAZING AND NONHAZING RETURNS. Testing the effects of hazing is not as simple as a laboratory exercise because there are no controls (Appendix IV). Returns within a basin can differ with time of release during a year and juvenile salmonids can linger in an estuary before or after hazing occurs, so it is not possible to compare hazed and nonhazed groups within a basin in the same year (Appendix IV). Because there is annual variation in returns for a basin and returns for a basin may differ from other basins, comparing results between a hazed and a nonhazed basin in the same year or between years with and without hazing at the same basin are also not rigorous (Appendix IV).
Nonetheless, such comparisons need to be done because of the interest in whether hazing may be an effective management tool in the recovery of salmonids and because hazing is supported by tax money. For example, proponents have stated that returns at the same basin (Nehalem) were greater for years with than without hazing and have sometimes suggested that returns for a hazed basin were greater than for nonhazed basin (Erickson 1989a,b,d; 1993, 1995a; Monroe 1995b, 1996a; Nokes 1995). Accordingly, comparisons are made at the same basin between years with and without hazing as well as between the Nehalem and nonhazed basins.
E-4b. DATA PRESENTATION. Graphical methods illustrate returns, abundance, and survival, so that the actual data can be examined to see if there are any consistent changes with hazing. Means are also often given because they can show whether returns are generally greater with hazing or not, and statistical analyses are used to determine if trends are significant. Together, these methods can help the reader evaluate the data with more insight than if only one type of data presentation is employed.
E-4c. STATISTICAL CHALLENGES. Tests are essential in determining if differences are statistically significant. However, there are challenges in using robust statistical methods in correlating returns with hazing because:
1) there are no controls to use in testing the survival of hazed vs.
nonhazed smolts (Appendix IV)
2) there is a delay in measuring the effects of hazing on abundance or
survival (section E-2), so that there are many variables other than
hazing that could influence the survival of smolts from hazed or
nonhazed estuaries (Appendix V)
3) the occurrence of hazing in 1985-1987 at Nehalem is uncertain
(section C-3a), so these years should not be included in analyses.
In spite of these challenges, it is still possible to examine the data cautiously to determine if hazing has had as much of a positive effect on returns as has been claimed.
E-4d. STATISTICAL HYPOTHESIS: HAZING >NONHAZING. Proponents have stated that hazing improves the survival of juvenile salmonids (Erickson 1989a,b,d; 1993, 1995a; Monroe 1995b, 1996a; Nokes 1995). Accordingly, their hypothesis is that the survival of salmon and steelhead is greater when hazing occurs than when it does not, and this is the one-tailed hypothesis that will be usually tested in this paper. A significance level of 0.05 has been chosen because it is most commonly used (Sokal and Rohlf 1981:164, Zar 1984:44, 98); if the probability is greater than 0.05, the result is considered statistically nonsignificant. However, if it appears that nonhazing returns may be significantly greater than for hazing, this difference is also tested.
E-4e. SAMPLE SIZE. Without an adequate sample size, it may not be possible to determine if a statistically significant difference exists (e.g., see "sample size" in Zar 1984:714). A sample size of five is sometimes considered to be a minimum, but this may not always be true. In this paper, I footnote tests with sample sizes of less than five.
There are three reasons for small sample sizes. First, I have only used data through 1998, and during that time there have only been four years of hazing (1988 and 1996-1998) at Tillamook Bay and three years (1996-1998) at Nestucca Bay; this is not a problem at the Nehalem, where hazing is known to have occurred for at least 10 years (1988-1994 and 1996-1998). Second, sample size is also reduced if there is a time delay in compiling fishery data; for example, 1994 fishery catch data became available in October 1998 (ODFW 1998b). Third, there can be a small sample size because adult steelhead and chinook return at various ages, so it may take several years before all returns for a given release of juvenile salmonids can be calculated (Appendix II-2d and II-3c). Although sample size is an issue, the one-tailed hypothesis that hazing has improved salmonid returns reduces its importance because if hazing is as beneficial as has been suggested, this should still show up with small sample sizes.
E-4f. STATISTICAL TESTS. The nonparametric Mann-Whitney test is used to determine if catches, returns, or survival are greater with or without hazing at Nehalem, Tillamook, or Nestucca Bays because it is unclear if the assumptions used for parametric tests are violated (e.g., data are normally distributed and have equal variances).
A paired Mann-Whitney test is used to compare the differences in catches, returns, or survival each year between the Nehalem basin and a basin without hazing. The underlying assumption to this test is that changes in factors other than hazing affect all basins the same. Although this assumption is not robust because basins often differ in yearly trends of returns (Jacobs and Cooney 1997, Chilcote 1998:25-26, 43, 45), survival (Lewis 1997:5-6), and catches (ODFW 1987, 1998b); it is appropriate to cautiously do this test because of the importance of the hazing issue and the lack of alternative tests.
A paired-sample test was also once used (section I-1a), when a Mann-Whitney test was not appropriate.
***************************************************************************
At Nehalem, "birds" became wary of the boat and would take off when the boat was about a mile away in 1996 (SPP 1996: April 17, April 21). In 1997, cormorants had become very wary of the boat by April 25 and scattered when the boat was a half mile or more away (SPP 1997).
At Tillamook, it is not reported how cormorants responded to hazing in 1996. The 1997 and 1998 SPP reports note if hazing was considered "successful" or not, but it is not explained what "successful" meant. Presumably, it meant that cormorants were chased from an area. The boat used for hazing could not chase cormorants from shallow waters at low tides, and once a hovercraft chased cormorants from such a situation at Kilchis Flats in 1997, but this did not seem to be a problem in 1998 (SPP 1997-1998).
At Nestucca, hazed "birds" would fly to another area to feed or roost on a sand bar, tree, or piling; sometimes they left the bay (SPP 1997-1998). It appears that some "birds" were often chased several times, and, on 11 May 1998, one hazer wrote: "Moved same birds all day."
Species other than cormorants identified as eating smolts in 1988 at Tillamook Bay included great blue herons, sea gulls, raccoons, mink, otters, and crows (Erickson 1995b). During 1996-1998, suspected predators included gulls at Nehalem (SPP 1996) and Nestucca Bays (SPP 1998), great blue herons at Nehalem (SPP 1996-1998) and Nestucca Bays (SPP 1998), Caspian terns at Nehalem (SPP 1996-1998) and Tillamook Bays (SPP 1997), green herons, crows, osprey, and mink at Nehalem (SPP 1996), and harbor seals at Tillamook (SPP 1997) and Nestucca Bays (SPP 1998).
Erickson (1989c, 1993, 1995b; Erickson in SPP 1996) considered cormorants to be the key smolt predator because he thought that they drove smolts up to the water surface or over to banks where other species could prey on smolts, and he indicated that if cormorants were driven away, then smolt predation by other predators would be insignificant. Jon Grigoraitis in SPP (1997) also thought that Caspian terns and great blue herons, but not harbor seals, were drawn to areas where cormorants had herded smolts, but he also reported that terns kept feeding after cormorants were chased away. Bob Rees at Tillamook Bay in 1997 (SPP 1997) noted that harbor seals were often not associated with cormorants. In 1999 at the Nehalem, Stahl et al. (2000:23) noted that double-crested cormorants, great blue herons, Caspian terns, and harbor seals sometimes foraged together and that cormorants seemed to attract other avian predators.
At Yaquina Bay, where I studied birds associated with hatchery releases of coho smolts (Bayer 1986), I often saw gulls, Caspian terns, and brown pelicans catching smolts independently of diving common murres (which were the predator of smolts of most concern there) or cormorants.
Animals other than cormorants were also disturbed during cormorant harassment. The cormorant hazing programs used cracker shells in 1988 and fast boats in later years (section C-2); this disturbance would not be specific to cormorants, especially if other animals were near the cormorants. For example, 25-150 waterfowl were often recorded as escaping the hazing at Nehalem during April-5 June 1997 (SPP 1997). Since black brant are particularly sensitive to human disturbance (e.g., Bayer 1996:743-744), they may have been disturbed by the cormorant hazing at Tillamook Bay, one of only three sites in Oregon where many overwinter (Bayer 1996), and where many are present in April when hazing occurs (Bayer 1989:44). Further, herons were noted as having been moved during cormorant hazing at Nestucca Bay (SPP 1998: April 24 [twice], May 23 [twice], and May 30), and for Nehalem in 1996, it is written: "Heron population is increasing and birds we saw very persistent in staying at Wheeler but got them to go out over state park" (SPP 1996: April 22). Hazers at Nestucca Bay may have sometimes attempted to chase wildlife other than cormorants; for example, the following comment was in SPP (1998:May 5): "3 loons can not move them."
On 27 April 1988, ADC personnel collected cormorants at Tillamook Bay that contained smolts (Hoffman and Hall 1988); this study has been used as evidence that cormorants were eating substantial numbers of smolts (Erickson 1988, 1995b; McAllister 1988). Unfortunately, this study was marred by faulty methods that make interpretation of their results tenuous (Bayer 1989:11, 25-26, 30).
During SPP (1996-1998), hazers sometimes observed cormorants eating smolts, and there were two reports of smolts recovered from cormorants. At Tillamook Bay on 28 April 1997, the contractor wrote that he had found three dead coho smolts in an area where he had hazed cormorants: two had their adipose fin clipped and were thus hatchery fish and one had all its fins intact and was thus probably a wild smolt (SPP 1997). Presumably, a cormorant had caught these smolts and regurgitated them when it was chased away. At Nestucca Bay on 23 April 1998, K. Delaney recovered and photographed five smolts that a cormorant regurgitated when it was chased (SPP 1998). Unfortunately, he did not identify the species of smolts or indicate if any had clipped fins and thus were hatchery fish, but one smolt (steelhead ?) was 8 inches (20 cm) and four smolts (coho ?) were 5-6 inches (13-15 cm)(SPP 1998).
In 1999 at the Nehalem, Stahl et al. (2000:28) reported that confirmed predators of radiotagged hatchery coho smolts were double- crested cormorants, harbor seals, and hooded mergansers; other species may have also preyed on these smolts. Prior to when hazing began on 1 April 1999, 30% of 10 radiotagged smolts were considered to have been preyed upon, and 55% of 20 smolts were thought to have been caught by predators after hazing began (Stahl et al. 2000:7, 28, 44).
In 1996 at Nehalem, cormorants started feeding on "pogeys" (sculpins ?) on May 3 and were reported to have switched over to feeding exclusively on "pink tailed perch babys" in the lower bay or "pogeys" in the upper bay around May 20 (SPP 1996). Alternate prey were not mentioned in SPP (1997), but, in 1998, cormorants were noted as feeding on nonsalmonid prey starting on June 9 (SPP 1998).
At Tillamook, the 1996 contractor observed that cormorants did not appear to be eating smolts after mid-May but were eating "sculpin, sandlance, shiners" (SPP 1996). In late May and June of 1997 and 1998, cormorants were noted as at least sometimes feeding on baitfish in the "lower estuary" (SPP 1997-1998).
At Nestucca, there were no notes of "birds" catching anything besides smolts.
From April through mid-June, there are about 13-15 hours of daylight between sunrise and sunset during which diurnal predators could prey on smolts. However, hazing does not appear to have occurred throughout the day at all three hazing sites (section C-5). At Nehalem, it is written that there was little cormorant activity after 3 PM in 1997, which is when the contractor stopped hazing (SPP 1997), and it is not expressed in the other reports if cormorant abundance decreased in the afternoon or increased again in the evening. At Tillamook, there does not appear to be enough counts throughout the day to determine if cormorant abundance changed during a day.
At Nestucca, the SPP (1996-1997) reports do not indicate changes in "bird" numbers during a day, but, in 1998, one of the three observers, J. Allen, recorded some details about his activities during six days from April 18 through May 3. His notes suggest that there were three periods of "bird" activity. Feeding "birds" were most abundant (usually 50- 150) and consistent during 5:45-8 AM PDT. Later in the morning at 10-12 AM PDT, as many as 50 "birds" were also sometimes active, and, in the evening from 4-7 PM PDT, there was often another period of activity by up to 65 "birds" that was more consistent than in late morning but usually with fewer "birds" than in the early morning. Throughout the rest of the day, he sometimes noted that there was "light activity," or he would note that "birds" were inactive in trees or on sand bars, snags, or pilings. Since he was present and chased "birds" throughout these days, his presence presumably inhibited "birds" from feeding more often or more continuously throughout the day, but cormorants elsewhere have also been observed to feed primarily in the morning (Whitfield and Blaber 1978) or to have 2-3 feeding peaks during a day (Bowmaker 1963:15, Bayer 1986:282).
Since cormorants were at least sometimes feeding in the early evening at Nestucca Bay during 1998, hazing at Nehalem that ceased at 3 PM PDT or that did not occur in the early evening at Tillamook Bay may have allowed some cormorants to feed then.
F-7a. INTRODUCTION. Because the methods of counting cormorants during SPP (1996-1998) are unclear (section D), it is speculative to interpret these counts and their accuracy. Nevertheless, these data are available, so they can be cautiously examined.
F-7b. NEHALEM. During 1996-1998, there was a similar pattern of high numbers of cormorants in April and much reduced abundance after about mid-May. In 1996, the high count was 400-500 "birds" (cormorants ?) on April 8 and 27, with 30 or less "birds" after May 20; apparently these counts were of "birds" at one location, and not the whole bay (SPP 1996). In 1997, the greatest number of cormorants seen in a single flock was 260, and no flocks with more than 100 cormorants were seen after May 15 (Table 5), but there were no counts of the whole bay. In 1998, a peak count of 640 cormorants in all of Nehalem Bay was noted in April, and 125 or fewer cormorants were found in May and early June (Table 5). This trend of a decline in cormorant abundance at the Nehalem after a week or so of hazing was also reported to have occurred during hazing prior to 1996 (Erickson 1995b).
In 1999, double-crested cormorant numbers were low in March and were much higher during April-May (Stahl et al. 2000:21, 49, 51).
---------------------------------------------------------------------------TABLE 5. Daily counts of cormorants at Nehalem Bay in 1997-1998 from SPP (1997-1998). 1997 and 1998 counts were done differently. 1997 counts are of the largest flock seen during a day at one location and are not a census of all Nehalem Bay; a count of 80+ cormorants was included in the 50-99 category, and a count of 100+ cormorants was placed in the 100-149 category. 1998 counts are reported to be the total seen at 7 or 8 AM in all of Nehalem Bay. .=zero (used to increase readability). MAX=maximum number of cormorants seen in a flock in 1997 or in Nehalem Bay estuary in 1998.
---------------------------------------------------------------------------
1997 No. of Days with a MAX Corm. Count in a Flock of... MAX
700 No. of
1- 11- 50- 100- 150- 200- 300- 400- 500- 600- or Corm-
Time Period 10 49 99 149 199 299 399 499 599 699 more orants
---------------------------------------------------------------------------
April 7-30 . 3 10 4 4 3 . . . . . 260
May 1-15 . 9 4 1 1 . . . . . . 165
May 16-31 7 8 1 . . . . . . . . 90
June 1-5 5 . . . . . . . . . . 4
---------------------------------------------------------------------------
---------------------------------------------------------------------------
1998 No. of Days with a Total Cormorant Count of....... MAX
700 No. of
1- 11- 50- 100- 150- 200- 300- 400- 500- 600- or Corm-
Time Period 10 49 99 149 199 299 399 499 599 699 more orants
---------------------------------------------------------------------------
April 24-30 . . . . 4 1 1 . . 1 . 640
May 1-15 . 10 4 1 . . . . . . . 125
May 16-31 . 7 7 2 . . . . . . . 125
June 1-15 1 12 2 . . . . . . . . 75
---------------------------------------------------------------------------
F-7c. TILLAMOOK. In 1996, peak counts of up to 850 cormorants were reported in April in the Wilson, Trask, and Tillamook Rivers and Tillamook Bay, but in May and June there were 100 or less (SPP 1996). However, it is unclear if these are rough estimates rather than accurate counts since exactly the same numbers are given for many days in a row (e.g., 100 each day during May 18-June 2)(SPP 1996).
1997-1998 counts were by a different contractor, and again it is not known how cormorants were censused. But it is indicated that the number of cormorants is the number at an "active site" (SPP 1997), so counts are not for the whole bay. One estimate of abundance is the maximum number of cormorants counted during a day, as shown in Table 6. No more than 250 cormorants were censused in 1997 and 1998, although generally fewer than 100 cormorants were recorded in both years (Table 6).
In 1999, double-crested cormorant numbers were low in March and were much higher during April-May (Stahl et al. 2000:21, 49, 51).
---------------------------------------------------------------------------TABLE 6. Daily peak counts of cormorants at portions of Tillamook Bay in 1997-1998. These counts are from SPP (1997-1998), were done by the same contractor, and were presumably done the same way; in 1997, counts were stated to be of the number of cormorants present at one site and thus not a census of all of Tillamook Bay. .=zero (used to increase readability). MAX=maximum number of cormorants seen at a site. *=approximately.
---------------------------------------------------------------------------
No. of Days with a MAX Cormorant Count at a Site of MAX
700 No. of
Time 0- 11- 50- 100- 150- 200- 300- 400- 500- 600- or Corm-
YR Period 10 49 99 149 199 299 399 499 599 699 more orants
---------------------------------------------------------------------------
97 April 8-30 1 9 6 1 . . . . . . . 95-105
97 May 1-15 6 8 . . . . . . . . . 36
97 May 16-31 3 4 2 1 1 1 . . . . . 200*
97 June 1-15 . 10 2 . . . . . . . . 60-70
98 April 21-30 . 2 . 1 1 1 . . . . . 180-220
98 May 1-15 1 4 1 1 2 5 . . . . . 200-250
98 May 16-31 4 9 1 . . . . . . . . 80
98 June 1-15 1 9 . . . 1 . . . . . 175-225
---------------------------------------------------------------------------
F-7d. NESTUCCA. Unfortunately, each report gives numbers of "birds" (which are presumably cormorants), and the methods of counting "birds" are not specified. The problem is compounded because there are three observers each year, and it is not stated if they counted "birds" the same way. However, counts appear to be of "birds" at a portion rather than all of Nestucca Bay.
The meaning of the "Number of Birds Observed" column in these reports is unclear and not defined. Generally, this column appears to include a sum of all "birds" counted at all sites, but at other times it may be the greatest number of "birds" seen at once in a portion of Nestucca Bay. Counts in this column for 1997 are given in Table 7; counts for 1996 and 1998 are not given because several numbers are given for some days and some counts include undefined symbols in SPP (1996), and one of the 1998 observers does not put numbers of "birds" in this column. Nevertheless, 1996 and 1998 "bird" numbers appear to be in the same order of magnitude as the numbers in SPP (1997) shown in Table 7.
In 1997, the greatest number (daily sum ?) of "birds" was reported in April and early May with a maximum of 350 "birds," and, in late May and early June, there was only one count (daily sum ?) greater than 100 "birds" (Table 7). Thus, the decrease in numbers in late May and early June is similar to that at Nehalem and Tillamook Bays, although the numbers given for Nestucca Bay can not be directly compared to those at the other bays because the Nestucca counts often appear to be a sum of all "birds" seen during a day, while counts at other bays represent the maximum seen at one area or a census of the whole estuary.
In SPP (1996-1998), the decline in the number or fishing activity of "birds" in June (1996-1997) or late May (1998) was attributed to an increase in recreational boat traffic.
In 1999, double-crested cormorant numbers were low in March and were much higher during April-May with an apparent peak in abundance during April 15-30 (Stahl et al. 2000:21, 49, 51).
---------------------------------------------------------------------------TABLE 7. Daily counts of "birds" (presumably cormorants) at Nestucca Bay in 1997. These data were in the "Number of Birds Observed" column in SPP (1997). Counts were made by three different observers, and it is unclear how each counted "birds" and if they did so the same way. At least some, perhaps most, counts appeared to be sums of the number seen during the day. Since "birds" were chased around the bay and may have been recounted several times each day, these numbers may overestimate the number of "birds" that were actually present. 1996 and 1998 counts in SPP (1996, 1998) are not given because they seem even more difficult to interpret. .=zero (used to increase readability). MAX=maximum number of cormorants reported.
---------------------------------------------------------------------------
No. of Days with a "Bird" Count of.................
700 MAX
Time 0- 11- 50- 100- 150- 200- 300- 400- 500- 600- or No. of
YR Period 10 49 99 149 199 299 399 499 599 699 more "Birds"
---------------------------------------------------------------------------
97 April 15-30 . 2 6 3 3 1 1 . . . . 350
97 May 1-15 . . 6 5 3 1 . . . . . 231
97 May 16-31 . 10 5 . 1 . . . . . . 170
97 June 1-15 1 8 6 . . . . . . . . 86
---------------------------------------------------------------------------
At the Nehalem, the presence of other potential smolt predators is mentioned in the 1996-1997 reports, but the method of counting these species is unclear. In SPP (1998), as many as 11 harbor seals were counted in late April at sites where "birds" were feeding, and seal abundance declined in May and particularly in June. "Terns" were the only other species counted, and they numbered 0-4 in late April, 0-20 in May, 0-2 during June 1-8, and 2-375 during June 9-15 (SPP 1998).
At Tillamook and Nestucca Bays, the presence of other predators was occasionally mentioned in SPP (1996-1998), but their abundance is not given.
For March-June 1999, Stahl et al. (2000:21-23, 50-51, 53, 67, 69) give censuses of some other potential predators at each bay. Caspian terns had brief peaks of about 50-200 birds at each site in mid-April, and seals regularly numbered 100-200 at the Nehalem but 50 or less at Tillamook and Nestucca Bays during the hazing period. There were also often 25-75 loons, grebes, or mergansers at Nehalem and Tillamook during hazing--other cormorant species and great blue herons were less numerous.
At Nehalem in 1996, cormorant abundance and/or feeding success was reported to be impaired during periods of muddy water in April, and cormorant abundance decreased during severe storms (SPP 1996). In contrast, cormorants foraged when the water was turbid in 1997, although the turbidity was less and briefer than in 1996 (SPP 1997); cormorants were also abundant during stormy weather in 1997 and 1998 (SPP 1997- 1998).
At Tillamook during the 24-27 April 1996 flood, no cormorants were found (SPP 1996), but there is no indication in the 1997-1998 reports about the effects, if any, of weather on cormorant activities. At the Nestucca, dirty water or flooding occurred in April 1996, but it is not clear if this affected the numbers or feeding of "birds" (SPP 1996).
Predation in the ocean beyond where hazers could go was not mentioned in the Nehalem or Tillamook SPP reports. But birds were noted in the ocean near the mouth of Nestucca Bay that were not hazed. On 11 May 1996, hundreds of cormorants were feeding outside the surf near the Nestucca at Pacific City (SPP 1996). In SPP (1998), J. Allen wrote that on April 18 there were 150 "birds" fishing in the jaws next to the surfline that could not be moved because of unsafe tide conditions, and D. Wenzinger observed that "birds" and seals were eating smolts outside of the first breaker on April 19.
***************************************************************************
The goal of the OCRSI, the Oregon Plan, and the ESA listing of coastal coho is the recovery of wild salmonids (OCSRI 1997a,b; NMFS 1998, Kitzhaber 1999). This goal is directed towards increasing the number of adults returning to spawn, but hazing may be more correlated with jack returns because jacks return sooner, so factors other than hazing may not affect their returns as much as for adults (Appendix V). Consequently, the returns of jack and adult coho, winter steelhead, and fall chinook to spawning grounds are examined; there are no spawning ground counts available for summer steelhead or summer and spring chinook.
At the Nehalem, Tillamook, and Nestucca basins, the average number of jacks per mile is less with hazing than without it, and jack abundance is not significantly greater with hazing (Table 8, Fig. 2). The differences between Nehalem and nonhazed stream jack returns are mixed. Graphical results suggest little if any change (Fig. 3). However, mean differences are greater with hazing at five of six basins, although annual differences are 1 jack/mi or less and are only statistically significant for two of six nonhazed streams (Table 8).
---------------------------------------------------------------------------TABLE 8. Statistical tests of whether the abundance of coho jacks at spawning areas is significantly greater with hazing. See Fig. 2 and Fig. 3 for details and yearly graphs. The larger mean is underlined. U=Mann-Whitney statistic for the hypothesis that the number of jacks at the Nehalem, Tillamook, or Nestucca basins or Nehalem-(basin) is more abundant with hazing than without it, N=number of years, P=probability, NS=not significant (one-tailed P>0.05).
---------------------------------------------------------------------------
Coho Jacks/Mi at Spawning Areas............ Hazing>
Hazed............... Nonhazed.............. Nonhazing
Comparison N Mean Range N Mean Range U P
---------------------------------------------------------------------------
Nehalem Basin 10 0.8 0-3.0 6 1.1 0.2-2.3 22.5 NSa
Tillamook Basin 4 0.1 0-0.4 15 0.5 0-1.4 13 NSa,b
Nestucca Basin 3 0.4 0-0.8 16 1.0 0-3.3 14.5 NSa,b
Nehalem-Necanicum 10 0 (-1.2)-1.7 5 -1.0 (-2.2)-(-0.3) 41.5 <0.05
Nehalem-Elk/Ecola 10 0 (-3.8)-1.1 5 -0.9 (-1.8)-0.4 40 <0.05
Nehalem-4th July 10 -1.4 (-2.7)-0.2 5 -1.4 (-2.9)-(-0.2) 27.5 NS
Nehalem-Yaquina 10 -0.6 (-2.5)-1.1 6 -1.1 (-3.1)-0.4 31 NS
Nehalem-Beaver 10 -1.7 (-5.5)-(-0.1) 6 -2.4 (-4.2)-(-1.1) 43.5 NS
Nehalem-Alsea 10 -0.9 (-3.2)-1.4 6 -1.0 (-1.7)-0 36 NS
a When the hypothesis that nonhazing abundance of jacks/mi is greater
than hazing is tested, the results are nonsignificant (one-tailed
P>0.05) for the Nehalem, Tillamook, and Nestucca Basins.
b The number of years with or without known hazing is less than five;
statistical tests would be more robust with larger sample sizes
(section E-4e).
---------------------------------------------------------------------------
---------------------------------------------------------------------------FIGURE 2. Peak number of coho jacks per mile during ODFW Spawning Fish Surveys at the Nehalem, Tillamook, and Nestucca Basins. Data are calculated from data through 1995 in Jacobs and Cooney (1997:Appendix II- D) and information available in August 1998 (1996-1997 data) and February 1999 (1997-1998 data) at http://osu.orst.edu/Dept/ODFW/other/spawn/data by summing the peak counts for each stream and dividing by the total miles surveyed in a basin. 1975-1979 data are not included because of missing data for Nehalem streams.
Jacks return to freshwater in the fall of the same calendar year that they entered the ocean as smolts, so hazing to protect migrating smolts in the spring of 1988 could affect the number of returning jacks during 1988 spawning surveys in November 1988 through January 1989.
At the Nehalem Basin, the total length of creeks surveyed was 5.14 mi in 1994 and 5.64 mi in other years and included the following (with the portion surveyed in parentheses): North Fork Cronin Creek (not surveyed in 1994; 0.5 mi in other years), West Humbug Creek (1.0 mi), Hamilton Creek (1.14 mi), Oak Ranch Creek (1.6 mi), and North Fork Wolf Creek (1.4 mi). Based on maps, all these creeks are tributaries of the Nehalem River and not of the North Fork of the Nehalem River.
At the Tillamook Basin, the total length of creeks surveyed was 3.6 mi in 1980, 5.2 mi in 1981-1985, 5.1 mi in 1986-1989 and 1992-1997, and 4.8 mi in 1990-1991 and included the following (with the portion surveyed in parentheses): Sam Downs Creek of the Kilchis River (1.0 mi), Cedar Creek of the Wilson River (2.9 mi in 1980-1985 and 2.8 mi thereafter), Upper Devil's Lake Fork of the Wilson River (0.7 mi), and Simmons Creek of Tillamook River (0.3 mi in 1990-1991 and 0.6 mi in other years).
At the Nestucca Basin, the total length of creeks surveyed was 1.6 mi in 1980 and 2.4 mi in other years and included the following (with the portion surveyed in parentheses): Clear Creek (0.8 mi, although in 1982 the lower 0.2 mi was dropped and 0.2 mi was added to the upper part of the survey), Bear Creek of the mainstem of the Nestucca River (not surveyed in 1980; 0.8 mi in other years), and Bear Creek of the Little Nestucca River (0.8 mi).
H=affected by hazing ?=unknown if Nehalem affected by hazing
o=not affected by hazing
------------------------ Coho Jacks/Mi ------------------------------------
Nehalem Basin Tillamook Basin Nestucca Basin
Nonhazing Mean= 1.1 Nonhazing Mean= 0.5 Nonhazing Mean= 1.0
Hazing Mean= 0.8 Hazing Mean= 0.1 Hazing Mean= 0.4
4.0-| -| -|
| | |
3.5-| -| -|
| | | o
J 3.0-| H -| -| o
a | H | | o
c 2.5-| H -| -| o
k | o H | | o
s 2.0-| o H -| -| o o
| o o H | |o oo o
p 1.5-| o o H -| -|o oo o
e | o o? H | o o |o oooo oo
r 1.0-|o-ooo?--HH-HH------ -|--o-o------------- -|o--oooo----oo------
|o ooo?? HH HH | oooo o o |ooooooo oo o H
M 0.5-|o ooo???HH HH oHHH -| oooooo Ho o -|oooooooo o oo o HH
i |ooooo???HHHHHH oHHH |ooooooooHoooo o H |oooooooo o oo o HH
0-|ooooo???HHHHHHHoHHH -|ooooooooHoooooooHHH -|ooooooooooooooooHHH
|''''|''''|''''|''' |''''|''''|''''|''' |''''|''''|''''|'''
80 85 90 95 80 85 90 95 80 85 90 95
Spawning Season
---------------------------------------------------------------------------FIGURE 3. Differences in peak numbers of coho jacks per mile during ODFW Spawning Fish Surveys between the Nehalem Basin and the Necanicum River, Elk/Ecola Creek, 4th of July Creek (Siletz River), Yaquina Basin, Beaver Creek, and Alsea Basin. These data are calculated from data through 1995 in Jacobs and Cooney (1997:Appendix II-D) and information available in August 1998 (1996-1997) and February 1999 (1997-1998 data) at http://osu.orst.edu/Dept/ODFW/other/spawn/data by summing the peak counts for each stream and dividing by the total miles surveyed in a basin. Hatchery fish have been excluded.
Cormorant harassment only occurred at the Nehalem and is known to have started there in the spring of 1988. Jacks return to freshwater in the fall of the same calendar year that they entered the ocean as smolts, so hazing to protect migrating smolts in the spring of 1988 could affect the number of returning jacks during 1988 spawning surveys in November 1988 through January 1989.
Basins along the north and central coasts north of the Siuslaw other than the Nehalem that were included are those without cormorant harassment, and they are arranged in the following graphs from north to south. The Nehalem Basin data are graphed and information about the creeks surveyed there are in Fig. 2. Other stream surveys include (with the portion surveyed in parentheses): the Upper Necanicum River (1.5 mi), the West Fork of Elk Creek/Ecola Creek (0.5 mi), Fourth of July Creek in the Siletz River Basin (0.8 mi), and North Fork Beaver Creek (1.0 mi). At the Yaquina Basin, the total length of streams surveyed was 2.4 mi in 1994 and 2.6 mi in other years and included the following (with the portion surveyed in parentheses): Salmon Creek (0.6 mi) and Upper Yaquina River (1.8 mi in 1994 and 2.0 mi in other years). At the Alsea Basin, the total length of streams surveyed was 2.3 mi in 1980, 4.86 mi in 1981, and 5.16 mi in other years and included the following (with the portion surveyed in parentheses): Horse Creek (1.0 mi), Nettle Creek (not surveyed in 1980 and 0.8 mi in other years), Lobster Creek (1.0 mi in 1981 and 1.3 mi in other years), Cherry Creek (not surveyed in 1980 and 0.76 mi in other years), and Wilson Creek (not surveyed in 1980 and 1.3 mi in other years).
H=Nehalem affected by hazing ?=unknown if Nehalem affected by hazing
o=not affected by hazing
------------------------ Coho Jacks/Mi ------------------------------------
Nehalem-Necanicum Nehalem-Elk Creek Nehalem-4th July
Nonhazing Mean= -1.0 Nonhazing Mean= -0.9 Nonhazing Mean= -1.4
J 3-| Hazing Mean= 0 -| Hazing Mean= 0 -| Hazing Mean= -1.4
a | | |
c 2-| -| -|
k | H | |
s 1-| H -| HH H -|
| H HH | HH H oHHH |
p 0-|=oooo???HHHHHHHoHHH====|=oooo???HHHHHHHoHHH====|=oooo???HHHHHHHoHHH
e | ooo?? H H HHo | o oo??? HH | ooo ???HHHHH oHHH
r -1-| o o?? H -| o oo??? HH -| o o ???HHHHH o H
| o o ? | o o??? H | o o ???HHHHH H
M -2-| o ? -| o o?? H -| o o ???HHH H
i | ? | ?? H | o o ? H H
-3-| ? -| ?? H -| o ?
| ? | ?? H | ?
-4-| -| ?? H -| ?
| | ?? |
-5-| -| ?? -|
| | ? |
-6-| -| ? -|
| | ? |
-7-| -| ? -|
| | |
|___________________ |___________________ |___________________
|''''|''''|''''|''' |''''|''''|''''|''' |''''|''''|''''|'''
80 85 90 95 80 85 90 95 80 85 90 95
Spawning Season
H=Nehalem affected by hazing ?=unknown if Nehalem affected by hazing
o=not affected by hazing
------------------------ Coho Jacks/Mi ------------------------------------
Nehalem-Yaquina Nehalem-Beaver Nehalem-Alsea
Nonhazing Mean= -1.1 Nonhazing Mean= -2.4 Nonhazing Mean= -1.0
J 3-| Hazing Mean= -0.6 -| Hazing Mean= -1.7 -| Hazing Mean= -0.9
a | | |
c 2-| -| -|
k | | | H
s 1-| H -| -| H
| o H | | H
p 0-|ooooo???HHHHHHHoHHH====|ooooo???HHHHHHHoHHH====|ooooo???HHHHHHHoHHH
e |oo ???H HHHHHo H |ooooo???HHH HHHoHHH |oooo ???H HHHHHo H
r -1-|oo ???H H Ho H -|ooooo???HHH HHHo HH -|oo o ???H HHH o
|oo ??? H Ho |ooo o??? HHHo HH |oo ? ?H HHH
M -2-|o ?? H Ho -|ooo ?? HHHo H -|o ? ?H H
i | ?? Ho |o o ?? Ho H | ? H H
-3-| ? o -|o o ?? H H -| ? H H
| | o ?? H | ?
-4-| -| o ?? H -|
| | ?? H |
-5-| -| ?? H -|
| | ?? H |
-6-| -| ?? H -|
| | ? |
-7-| -| ? -|
| | |
|___________________ |___________________ |___________________
|''''|''''|''''|''' |''''|''''|''''|''' |''''|''''|''''|'''
80 85 90 95 80 85 90 95 80 85 90 95
Spawning Season
---------------------------------------------------------------------------
The number of coho adults per mile has generally decreased since hazing began at the Nehalem, Tillamook, and Nestucca basins (Fig. 4), and average returns are greater without hazing than with it (Table 9). These differences are not significant at the Nehalem and Tillamook, but returns are significantly greater without hazing at the Nestucca, although there are only two years of hazing-affected returns (Table 9: footnote a). The averages of 1-7 adults/mi for hazing-affected counts (Table 9) are far below the PFMC goal of 42 adults/mi (PFMC 1996:III-5).
Graphs suggest little change in differences between Nehalem and nonhazed streams in adult returns (Fig. 5). But mean differences are greater with hazing for four of six streams, although the increase in averages with hazing is small (less than 4 adults/mi)(Table 9), and the differences are not significantly greater with hazing (Table 9).
---------------------------------------------------------------------------TABLE 9. Statistical tests of whether the abundance of coho adults at spawning areas is significantly greater with hazing. See Fig. 4 and Fig. 5 for details and yearly graphs. The larger mean is underlined. U=Mann-Whitney statistic for the hypothesis that the number of adults at the Nehalem, Tillamook, or Nestucca basins or Nehalem-(basin) is more abundant with hazing than without it, N=number of years, P=probability, NS=not significant (one-tailed P>0.05).
---------------------------------------------------------------------------
Wild Coho Adults/Mi at Spawning Areas...... Hazing>
Hazed............... Nonhazed.............. Nonhazing
Comparison N Mean Range N Mean Range U P
---------------------------------------------------------------------------
Nehalem Basin 9 6.6 2.5-18.3 7 9.7 2.0-24.8 23 NSa
Tillamook Basin 3 2.7 1.2-5.1 16 5.3 0.8-11.9 14.5 NSa,b
Nestucca Basin 2 1.0 0-2.1 17 7.8 1.3-17.9 1 NSa,b
Nehalem-Necanicum 9 1.6 (-2.9)-8.7 6 -2.0 (-20.1)-15.5 32 NS
Nehalem-Elk/Ecola 9 -0.9 (-11.5)-6.3 6 -4.4 (-12.6)-6.8 36 NS
Nehalem-4th July 9 -12.0 (-31.3)-1.0 6 -11.8 (-42.6)-7.3 27 NSa
Nehalem-Yaquina 9 -10.0 (-49.4)-4.0 7 -5.8 (-28.9)-4.1 23 NSa
Nehalem-Beaver 9 -10.8 (-48.2)-5.3 7 -13.3 (-36.1)-0.8 37 NS
Nehalem-Alsea 9 -5.4 (-24.1)-8.7 7 -8.2 (-31.0)-2.7 39 NS
a When the hypothesis that nonhazing abundance of adults/mi is greater than
hazing is tested, the results are nonsignificant (one-tailed P>0.05),
except for the Nestucca Basin (U=33, one-tailed P<0.05), where there
were only two years of hazing data.
b The number of years with or without known hazing is less than five;
statistical tests would be more robust with larger sample sizes
(section E-4e).
---------------------------------------------------------------------------
FIGURE 4. Peak number of wild adult coho per mile during ODFW Spawning Fish Surveys at the Nehalem, Tillamook, and Nestucca Basins. Data are calculated from data through 1995 in Jacobs and Cooney (1997:Appendix II-D) and information available in August 1998 (1996-1997) and February 1999 (1997-1998 data) at http://osu.orst.edu/Dept/ODFW/other/spawn/data by summing the peak counts for each stream and dividing by the total miles surveyed in a basin. Hatchery fish have been excluded. See the legend
of Fig. 2 for details about the streams surveyed. 1975-1979 data are not included because of missing data for Nehalem streams.
Adults return to freshwater in the calendar year after they entered the ocean as smolts, so hazing to protect migrating smolts in the spring of 1988 could affect the number of returning adults during 1989 spawning surveys in November 1989 through January 1990.
H=affected by hazing ?=unknown if Nehalem affected by hazing
o=not affected by hazing
-------------------------- Wild Coho Adults/Mi ----------------------------
Nehalem Basin Tillamook Basin Nestucca Basin
Nonhazing Mean= 9.7 Nonhazing Mean= 5.3 Nonhazing Mean= 7.8
Hazing Mean= 6.6 Hazing Mean= 2.7 Hazing Mean= 1.0
30-| -| -|
A | | |
d 25-| o -| -|
u | o | |
l 20-| o -| -|
t | o H H | | o
s 15-| o ? H H -| -| oo
| o o ? H H | o | ooooo
p 10-|o-o-oo?--H-H------------|--o-o-------------------|o-ooooo----o-------
e |o o oo? H H | oo ooo o |ooooooo o o
r 5-|ooo oo???H HHH -|ooo oooooH o oo -|ooooooo o o ooo
|oooooo???HHHHHHHoHH |oooooooooH o oooH |oooooooooooooooo H
M 0-|oooooo???HHHHHHHoHH -|oooooooooHoooooooHH -|oooooooooooooooooHH
i |''''|''''|''''|''' |''''|''''|''''|''' |''''|''''|''''|'''
80 85 90 95 80 85 90 95 80 85 90 95
Spawning Season
---------------------------------------------------------------------------
FIGURE 5. Differences in peak numbers of wild adult coho per mile during ODFW Spawning Fish Surveys between the Nehalem Basin and the Necanicum River, Elk/Ecola Creek, 4th of July Creek (Siletz River), Yaquina Basin,
Beaver Creek, and Alsea Basin. These data are calculated from data
through 1995 in Jacobs and Cooney (1997:Appendix II-D) and information
available in August 1998 (1996-1997) and February 1999 (1997-1998 data)
at http://osu.orst.edu/Dept/ODFW/other/spawn/data by summing the peak counts for each stream and dividing by the total miles surveyed in a basin. Hatchery fish have been excluded.
Cormorant harassment only occurred at the Nehalem and is known to have started there in the spring of 1988. Adults return to freshwater in the calendar year after they entered the ocean as smolts, so hazing to protect migrating smolts in the spring of 1988 could affect the number of returning adults during 1989 spawning surveys in November 1989 through January 1990.
See the Legends of Fig. 2 and Fig. 3 for details about the streams surveyed.
H=Nehalem affected by hazing ?=unknown if Nehalem affected by hazing
o=not affected by hazing
------------------------------ Wild Coho Adults/Mi ------------------------
Nehalem-Necanicum Nehalem-Elk Creek Nehalem-4th July
Nonhazing Mean= -2.0 Nonhazing Mean= -4.4 Nonhazing Mean= -11.8
A 30-| Hazing Mean= 1.6 -| Hazing Mean= -0.9 -| Hazing Mean= -12.0
d | | |
u 20-| -| -|
l | o | |
t 10-| o H H -| -|
s | o ? H H | o H HH H | o
0-|=ooooo???HHHHHHHoHH====|=ooooo???HHHHHHHoHH====|=ooooo???HHHHHHHoHH
p | oo ?? Ho | o o o HH H oH | ooo o???HH H HHoHH
e -10-| o ? o -| o o oH -| o o o???H H HH H
r | o | o | o o? ?H H H
-20-| o -| -| o? H H H
M | | | o? H H H
i -30-| -| -| o? H
| | | o
-40-| -| -| o
| | | o
|___________________ |___________________ |___________________
|''''|''''|''''|''' |''''|''''|''''|''' |''''|''''|''''|'''
80 85 90 95 80 85 90 95 80 85 90 95
Nehalem-Yaquina Nehalem-Beaver Cr Nehalem-Alsea
Nonhazing Mean= -5.8 Nonhazing Mean= -13.3 Nonhazing Mean= -8.2
A 30-| Hazing Mean= -10.0 -| Hazing Mean= -10.8 -| Hazing Mean= -5.4
d | | |
u 20-| -| -|
l | | |
t 10-| o -| -| o H
s | o o H | H | o o H
0-|oooooo???HHHHHHHoHH====|oooooo???HHHHHHHoHH====|oooooo???HHHHHHHoHH
p |oo o??? H HHHHoH |oooo o? HHHHoHH |oo o o???HH HH oHH
e -10-|o ? HHHo -|o o o? H HHoHH -| o o??? H o H
r |o ? HHo |o o? H oHH | o o ? H
-20-|o Ho -|o o? H -| o ? H
M | Ho | o? H | o ? H
i -30-| Ho -| o H -| o
| H | o H |
-40-| H -| H -|
| H | H |
-50-| H -| H -|
|___________________ |___________________ |___________________
|''''|''''|''''|''' |''''|''''|''''|''' |''''|''''|''''|'''
80 85 90 95 80 85 90 95 80 85 90 95
Spawning Season
---------------------------------------------------------------------------
Winter steelhead spawning surveys commenced in the spring of 1998 in portions of many Tillamook County streams (see http://osu.orst.edu/Dept/ODFW/other/spawn/data). Previously, the only stream with consistent surveys along the north and central coast of Oregon is the Salmonberry River in the Nehalem Basin (Chilcote 1998:40; Bob Buckman, ODFW, pers. comm.).
The abundance of winter steelhead at the Salmonberry River has not increased after hazing commenced and has not reached Chilcote's (1998:45) recommended equilibrium level to sustain the population in 4 of 7 years that were affected by hazing (Fig. 6). Indeed, the estimated average abundance is lower with hazing (Fig. 6), and adult returns during 1991-1996 that were hazing-affected are not significantly greater than for the nonhazing-affected years of 1975-1986 (Mann-Whitney U=22.5; N1=6, N2=12, one-tailed P>0.10).
---------------------------------------------------------------------------FIGURE 6. Number of spawning winter steelhead per mile at the Salmonberry River, a tributary of the Nehalem River. These ODFW data are estimated from Chilcote's (1998:45) graph. Surveys were during April and May, so these are thought to be primarily wild steelhead because hatchery steelhead typically spawn in January and February (Chilcote 1998:45). The estimated natural equilibrium level for the Salmonberry River is 31 spawners per mile (Chilcote 1998:45). In 1992, fishing regulations were changed to require the release of wild steelhead (Chilcote 1998:31), so it could be expected that the number of spawning steelhead would consequently increase.
Since naturally spawned smolts typically spend two years in freshwater, smolts originating from the 1986 Spawning Year (Brood Year) would have migrated to the ocean during 1988 and returned as adults in the 1990 Spawning Year as 2-salts and in the 1991 Spawning Year as 3-salts with some repeat spawners returning in later years (Appendix II-2d).
Salmonberry River Winter Steelhead Spawners/Mile
H=affected by hazing ?=unknown if affected by hazing
O=not affected by hazing --=equilibrium level
P=only 2-salts affected by hazing
#=fishers required to release wild steelhead
S 120-| O Nonhazing 1975-1986 Mean= 42/mi
t -| O Hazing 1991-1996 Mean= 26/mi
e 110-| O
e -| O
l 100-| O
h -| O
e 90-| O
a -| O
d 80-| O
-| O ?
S 70-| O O ?
p -| O O ?
a 60-| O O ?
w -| O O O ?
n 50-| O O O O O ? ?
e -| O O O O O O ? ? H
r 40-| O O O O O O ? ? H H
s -| O O O O O O O ? ? H H
30-|-O--O--O--O--O--O--------------O--O--?--?--?--P--H-----------H---
p -| O O O O O O O O O ? ? ? P H H H H
e 20-| O O O O O O O O O O ? ? ? P H H H H
r -| O O O O O O O O O O ? ? ? P H H H H H
10-| O O O O O O O O O O O ? ? ? P H H H H H
M -| O O O O O O O O O O O O ? ? ? P H H H H H H
i 0-| O O O O O O O O O O O O ? ? ? P H H H H H H
l |____________________________________________________#__#__#__#__#
e | | | | | | | | | | | | | | | | | | | | | |
75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96
Spawning Year
---------------------------------------------------------------------------
---------------------------------------------------------------------------TABLE 10. Statistical tests of whether the abundance of fall chinook jacks at spawning grounds is significantly greater without hazing. The hypothesis throughout the rest of this paper is that abundance is greater with than without hazing, but the consistently greater means for nonhazing years make it clear that this was obviously not true for fall chinook, so a more meaningful test is if nonhazing abundance is greater than hazing. See Fig. 7 and Fig. 8 for details and yearly graphs of these data. The larger mean is underlined. U=Mann-Whitney statistic for the hypothesis that the number of jacks at the Nehalem, Tillamook, or Nestucca basins or Nehalem-(basin) is more abundant without hazing, N=number of years, P=probability, NS=not significant (one-tailed P>0.05).
---------------------------------------------------------------------------
Fall Chinook Jacks/Mi at Spawning Areas.... Nonhazing>
Hazed.............. Nonhazed.............. Hazing
Comparison N Mean Range N Mean Range U P
---------------------------------------------------------------------------
Nehalem (Humbug) 9 2.0 0-5.0 12 15.1 1.0-40.0 90 <0.01
Tillamook Basin 3 1.1 0.4-2.3 15 2.9 0.5-5.9 36 NSa
Nestucca Basin 2 0.4 0-0.8 18 4.7 0-17.0 32.5 <0.05a
Nehalem-Siletz 9 0.6 (-5.0)-3.3 11 11.3 (-3.8)-37.5 74 <0.05
Nehalem-Yaquina 9 0.3 (-3.5)-2.8 7 11.2 (-0.5)-31.9 50 <0.05
Nehalem-Alsea 9 -1.8 (-5.0)-2.0 12 11.3 (-8.0)-39.0 90.5 <0.01
a The number of years with or without known hazing is less than five;
statistical tests would be more robust with larger sample sizes
(section E-4e).
---------------------------------------------------------------------------
---------------------------------------------------------------------------
FIGURE 7. Peak number of fall chinook jacks per mile during ODFW Spawning Fish Surveys at the Nehalem, Tillamook, and Nestucca Basins.
These returns are calculated from data through 1995 in Jacobs and Cooney
(1997:Appendix I-B) and information available in August 1998 (1996-1997
data) and February 1999 (1997-1998 data) at
http://osu.orst.edu/Dept/ODFW/other/spawn/data by summing the peak counts
for each stream and dividing by the total miles surveyed in a basin. It
is unclear if some hatchery fish may have been included (see Jacobs and
Cooney 1997:6-8).
The only Nehalem Basin stream surveyed since 1975 was 1.0 mi of Humbug Creek. At the Tillamook Basin, the total length of streams surveyed since 1981 was 5.6 mi and included the following (with the portion surveyed in parentheses): Little South Fork of the Kilchis River (1.0 mi), Cedar Creek of the Wilson River (2.8 mi), and Tillamook River (1.8 mi). At the Nestucca Basin, the total length of streams surveyed since 1980 was 1.2 mi and included (with the portion surveyed in parentheses): Clear Creek (0.8 mi) and Niagara Creek (0.4 mi). Other streams in the Tillamook and Nestucca Basins were excluded if surveys commenced in 1986, had missing data, or changed survey length.
H=affected by hazing ?=unknown if Nehalem affected by hazing
o=not affected by hazing
----------------- Fall Chinook Jacks/Mi------------------------------------
Nehalem Tillamook
Nonhazing Mean= 15.1 Nonhazing Mean= 2.9
Hazing Mean= 2.0 Hazing Mean= 1.1
39-o 31-oo-40
28-o| ||
|| ||
||| || |
16-||| || -|
||| || |
14-||| || -|
|oo o oo |
12-|ooo o oo -|
J |ooo o oo |
a 10-|ooo o oo -|
c |ooo o oo |
k 8-|ooo o oo? -|
s |ooo o oo? |
6-|oooo o oo?? -| o o
p |oooo o oo?? H | o ooo
e 4-|oooo oooo??? H H -| o ooo o
r |oooo o oooo??? H HH | o ooo ooo
2-|oooooo oooo???H H HHo H -| ooooooooHooo
M |ooooooooooo???HH H HHoHH | ooooooooHoooooooH
i 0-|ooooooooooo???HHHHHHHoHH -|______ooooooooHoooooooHH
|''''|''''|''''|''''|''' |''''|''''|''''|''''|'''
75 80 85 90 95 75 80 85 90 95
Spawning Season
Nestucca
Nonhazing Mean= 4.7
Hazing Mean= 0.4
| o
16-| o
| o
14-| o
| o
12-| o
J | o
a 10-| o oo o
c | o oo o
k 8-| o oo o
s | oo oo o
6-| ooooo o o
p | ooooo o o
e 4-| o ooooo o o
r | o ooooo ooooo
2-| o ooooo ooooooo
M | ooooooooooooooooo H
i 0-|____ooooooooooooooooooHH
|''''|''''|''''|''''|'''
75 80 85 90 95
Spawning Season
---------------------------------------------------------------------------
FIGURE 8. Differences in peak numbers of fall chinook jacks per mile during ODFW Spawning Fish Surveys between the Nehalem Basin (Humbug Creek) and Sunshine Creek (Siletz River), Yaquina Basin, and Buck Creek
(Alsea River). These returns are calculated from data through 1995 in
Jacobs and Cooney (1997:Appendix II-D) and information available in
August 1998 (1996-1997 data) and February 1999 (1997-1998 data) at
http://osu.orst.edu/Dept/ODFW/other/spawn/data by summing the peak counts for each stream and dividing by the total miles surveyed in a basin. It is unclear if some hatchery fish may have been included (see Jacobs and Cooney 1997:6-8).
Basins along the north and central coasts north of the Siuslaw other than the Nehalem that were included are those without cormorant harassment. These basins are arranged in the following graphs from north to south. The Nehalem Basin (Humbug Creek, 1.0 mi) data are graphed in Fig. 7. The total length of streams surveyed was 1.2 mi at Sunshine Creek in the Siletz Basin and 1.0 mi at Buck Creek in the Alsea Basin. At the Yaquina Basin, the total was 2.6 mi (except in 1994 when the total was 2.35 mi) and surveys were in the Upper Yaquina River (2.0 mi, except in 1994 when 1.75 mi was surveyed) and Salmon Creek (0.6 mi). Other streams in these basins were excluded if surveys commenced in 1986 or had missing data.
H=Nehalem affected by hazing ?=unknown if Nehalem affected by hazing
o=not affected by hazing
----------------------- Fall Chinook Jacks/Mi -----------------------------
Nehalem-Siletz Nehalem-Yaquina
Nonhazing Mean= 11.3 Nonhazing Mean= 11.2
Hazing Mean= 0.5 Hazing Mean= 0.3
o-38
29-o 29-o| 29-oo-32
| || ||
16-| o oo -| oo
| o oo | oo
14-| o oo -| oo
| o oo | o oo
12-| o o oo -| o oo
| o o oo | o oo
J 10-| oo o oo -| o oo
a | oo o oo | o oo
c 8-| oo o oo -| o oo
k | oo o oo | o oo
s 6-| ooo o oo ? -| o oo
| ooo o oo ? | o oo
p 4-| ooo o oo?? -| o oo ?
e | ooo o oo??? H HH | o o oo ?? H
r 2-| ooo oooo??? H HHo -| o oooo ??H H
| ooo o oooo??? H HHoH | o oooo???HH H H H
M 0-| oooooooooo???HHHHHHHoHH -| oooooo???HHHHHHHoHH
i | o o H H H | o H H
-2-| o H -| H
| o H | H
-4-| o H -| H
| H |
-6-| -|
| |
-8-| -|
|________________________ |________________________
|''''|''''|''''|''''|''' |''''|''''|''''|''''|'''
75 80 85 90 95 75 80 85 90 95
Spawning Season
Nehalem-Alsea
Nonhazing Mean= 11.3
Hazing Mean= -1.8
o-39
27-o| 29-oo-25
|| ||
16-|oo oo
|oo oo
14-|oo oo
|oo oo
12-|oo oo
|oo oo
J 10-|ooo oo
a |ooo oo
c 8-|ooo oo
k |ooo oo
s 6-|oooo oo
|oooo oo ?
p 4-|oooo oooo ?
e |oooo oooo ?
r 2-|oooo oooo ? H
|oooo o oooo ? H H
M 0-|ooooooooooo???HHHHHHHoHH
i | o o ? ?HHHHH H H
-2-| o o ?HHHH H
| o ?HH H
-4-| o ? H H
| o ? H H
-6-| o ?
| o ?
-8-| o
|________________________
|''''|''''|''''|''''|'''
75 80 85 90 95
Spawning Season
---------------------------------------------------------------------------
At the Nehalem, the abundance of wild fall chinook adults shows a significant decline with hazing (Fig. 9, Table 11). Comparisons of adult returns to the Tillamook and Nestucca Basins are not appropriate because there have not yet been enough consecutive years of hazing, so that all age classes would be affected. Relative to nonhazed streams, adult returns affected by hazing at the Nehalem are also significantly less (Table 11, Fig. 10).
The decline with hazing may be related to increased chinook fishery catches (section J-3).
---------------------------------------------------------------------------TABLE 11. Statistical tests of whether the abundance of fall chinook adults at spawning areas is significantly greater with no hazing. The hypothesis throughout the rest of this paper is that hazing abundance is greater than for nonhazing, but the consistently greater means for nonhazing make it clear that this is obviously not true for fall chinook, so a more meaningful test is if nonhazing abundance is greater than hazing. See Fig. 9 and Fig. 10 for details and yearly graphs of these data. The larger mean is underlined. U=Mann-Whitney statistic for hypothesis that the number of adults at the Nehalem, Tillamook, or Nestucca basins or Nehalem-(basin) is more abundant with no hazing than with it, N=number of years, P=probability, NS=not significant (one-tailed P>0.05).
---------------------------------------------------------------------------
Fall Chinook Adults/Mi at Spawning Areas... Nonhazing>
Hazed.............. Nonhazed.............. Hazing
Comparison N Mean Range N Mean Range U P
---------------------------------------------------------------------------
Nehalem (Humbug) 4 69.0 50.0-86.0 12 142.8 61.0-280.0 42 <0.05a
Nehalem-Siletz 4 8.0 (-15.7)-35.8 11 107.3 36.7-240.8 44 <0.01a
Nehalem-Yaquina 4 -4.7 (-55.7)-32.7 7 111.9 28.4-231.9 27 <0.05a
Nehalem-Alsea 4 23.5 (-2.0)-37.0 12 108.5 16.0-225.0 44 <0.01a
a The number of years with all 3-6 yr olds affected by hazing is less
than five; statistical tests would be more robust with larger
sample sizes (section E-4e).
---------------------------------------------------------------------------
FIGURE 9. Peak number of fall chinook adults per mile during ODFW Spawning Fish Surveys at the Nehalem Basin. Data are calculated from data through 1995 in Jacobs and Cooney (1997:Appendix I-B) and information available in August 1998 (1996-1997 data) and February 1999 (1997-1998 data) at http://osu.orst.edu/Dept/ODFW/other/spawn/data by summing the peak counts for each stream and dividing by the total miles surveyed in a basin. It is unclear if some hatchery fish may have been included (see Jacobs and Cooney 1997:6-8). The only Nehalem Basin stream surveyed since 1975 was 1.0 mi of Humbug Creek.
H=all affected by hazing ?=unknown if affected by hazing
o=not affected by hazing
P=part of, but not all, 3-6 yr old adults affected by hazing
Nehalem Fall Chinook Adults/Mi
300-| Nonhazing 1975-86 Mean= 142.8
A -| o Hazing 1993-96 Mean= 69.0
d 250-| o o
u -| o o ?
l 200-| o o ?
t -| o o o o ?
s 150-| o o o o o o ? ? ? P
-| o o o o o o o ? ? ? P
p 100-| o o o o o o o o o ? ? ? P P P
e -| o o o o o o o o o o o ? ? ? P H H P P
r 50-|-o-o-o-o-o-o-o-o-o-o-o-o-?-?-?-P-P-P-H-H-H-H-P-P-- 50
-| o o o o o o o o o o o o ? ? ? P P P H H H H P P
M 0-| o o o o o o o o o o o o ? ? ? P P P H H H H P P
i | ' ' ' ' | ' ' ' ' | ' ' ' ' | ' ' ' ' | ' ' '
75 80 85 90 95
Spawning Season
---------------------------------------------------------------------------
FIGURE 10. Differences in peak numbers of fall chinook adults per mile during ODFW Spawning Fish Surveys between the Nehalem Basin (Humbug Creek) and Sunshine Creek (Siletz River), Yaquina Basin, and Buck Creek (Alsea River). These data are calculated from data through 1995 in Jacobs and Cooney (1997:Appendix II-D) and information available in August 1998 (1996-1997 data) and February 1999 (1997-1998 data) at http://osu.orst.edu/Dept/ODFW/other/spawn/data by summing the peak counts for each stream and dividing by the total miles surveyed in a basin. It is unclear if some hatchery fish may have been included (see Jacobs and Cooney 1997:6-8).
Basins along the north and central coasts north of the Siuslaw other than the Nehalem that were included are those without cormorant harassment. These basins are arranged in the following graphs from north to south. The Nehalem Basin (Humbug Creek, 1.0 mi) data are graphed in Fig. 9. The total length of streams surveyed was 1.2 mi at Sunshine Creek in the Siletz Basin and 1.0 mi at Buck Creek in the Alsea Basin. At the Yaquina Basin, the total was 2.6 mi (except in 1994 when the total was 2.35 mi) and surveys were in the Upper Yaquina River (2.0 mi, except in 1994 when 1.75 mi was surveyed) and Salmon Creek (0.6 mi). Other streams in these basins were excluded if surveys commenced in 1986 or had missing data.
H=Nehalem affected by hazing ?=unknown if Nehalem affected by hazing
o=not affected by hazing
P=part of, but not all, 3-6 yr old adults affected by hazing
------------------------ Fall Chinook Adults/Mi ---------------------------
Nehalem-Siletz Nehalem-Yaquina
Nonhazing Mean= 107.3 Nonhazing Mean= 111.9
Hazing Mean= 8.0 Hazing Mean= -4.7
A 250-| o -|
d | o | oo
u 200-| o ? -| oo
l | oo ? | oo ?
t 150-| o oo ? -| oo ? ?
s | ooo oo ? | o oo ??? P
100-| oooo o oo ? ? P -|